Apparently, at least one person has misinterpreted my mail as an attack - it
was not meant to be. I was trying to get all of us closer to having RENDER
supported in other Xservers, so that more application writers will use it, and
more users can benefit from it. I wouldn't be doing this if I thought that
"RENDER should die", and I apologize to anyone who misunderstood me.
> I am unaware of Sun's contribution to XFree86 up to this point.
I don't know what this has to do with anything, but a quick find | xargs grep
of the XFree86 4.2.0 source tree shows 445 files containing "Sun Microsystems."
This is admittedly a small portion of the source tree, but Sun has contributed
for many years to the X Consortium/X.org distributions that XFree86 is based
on. Most recently, we open sourced quite a bit of i18n code that I believe
XFree86 has adopted. We also continue to contribute financial support to
X.org, along with all the other X.org members, to continue the maintance of
the X.org open source distribution for the benefit of everyone, including
XFree86.
> Could it be that
> technology is changing and thus backwards compatiblity is something that
> will hamper future progress.
You may not need backwards compatibility, but for the vast majority of computer
users, both commercial and personal, a computer that won't run their software
isn't very useful. Not everyone has the ability, resources, or desire to
rewrite software to make it compatible with every change that comes along.
As Keith already answered, and the XFree86 10th anniversary messages have also
pointed out, backwards compatibility is just as important to XFree86 as it is
to Sun.
Fortunately, X provides us mechanisms to build in both backwards compatibility
and future extensibility. I am very glad to find out that Keith has done a
good job in keeping both in mind with his RENDER work - the primary concern I
had was this was not visible from the available documentation.
> I think you are asking too much and frankly giving very little in return
> except the promise that RENDER
> and XFREE86 should die a slow and painful death to make it easy for SUN
> engineers.
Neither I nor anyone else at Sun wants to see RENDER or XFree86 die. That is
just about the worst possible outcome I can imagine, and fortunately, one I
think is almost uncertain to occur anytime soon. In fact, a strong XFree86
is very beneficial to Sun's desires for the future of the computing market.
The most likely outcome of this discussion, and the one that I think everyone
wants, is that we end up with a document that server writers & application
writers can use to determine what portions of RENDER functionality are
guaranteed to be supported by servers that claim to provide a certain version
of the RENDER extension. Then when GNOME, KDE, Mozilla, etc. query the RENDER
extension, they know what requests they can make given the extension version
returned, no matter what server is running, whether it be XFree86, Xsun, or any
other server. This is vital not only to spreading the RENDER extension to other
servers, but in managing the future evolution of RENDER in XFree86 as well.
BTW, just to be clear, I am just an X Engineer at Sun, not an official
spokesperson. Nothing I say should ever be taken too seriously, and especially
should not be taken as the official position of Sun.
--
-Alan Coopersmith- [EMAIL PROTECTED]
Sun Microsystems, Inc. - Software Systems Group
Cust. Advocacy & Tech Services: X11 Engineering
_______________________________________________
Render mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/render
